The Bias in the Solution to Bias
If you are interested in healthcare and have been following up with the news lately, you likely have encountered this newly published study by researchers at Hofstra School of Medicine and Harvard Medical School (http://www.healio.com/psychiatry/pediatrics/news/online/%7B847fb66a-ae9e-4d1d-9a73-2ba274af31c4%7D/analysis-shows-large-racial-disparities-in-mental-health-care-for-children-young-adults). The researchers found that, when having mental health or substance abuse problems, African American and Hispanic children are more likely to be punished or incarcerated, instead of getting the care they need. The non-Hispanic White counterparts, on the other hand, are far more likely to receive treatment. In a nutshell, our healthcare system is not impartial. A given individual’s access to healthcare services could depend on so many things, including race and ethnicity, geographical region, and socioeconomic status. On top of that, even when the patient is in a doctor’s office, the quality of care they receive could vary. As you could probably tell, the disparities of the access to and quality of healthcare services are quite concerning. Thankfully, many scholars and policymakers have started to intervene. But are these interventions really doing what we expect them to do?
Before diving into the broad array of interventions, let’s start with the pay-for-performance model. Pay-for-performance model, or P4P, is a type of performance incentive program. Wikipedia defines it as,
“a payment model that offers financial incentives to physicians, hospitals, medical groups, and other healthcare providers for meeting certain performance measures.”
The idea is simple. The quality of care might not be satisfying all the time. The pay-for-performance system acts as teacher giving out red stars to the most well behaved kids in kindergarten. It is meant to reward those that put efforts into their services, and ultimately improve the quality of care.
As almost all of solutions that people come up with, the pay-for-performance model is not perfect. In fact, it is a system that is easily abused. This might become more clear when you think about racial and ethnic disparities, and the reasons behind the subpar quality of performance in the healthcare industry. Due to the cultural barrier to their physicians, racial and ethnic minority patients are intrinsically more difficult to provide high quality care for. And the traditional pay-for-performance model rewards whoever reaches a set level of performance, as opposed to the greatest level of improvement. In other words, the system sets a bar. As long as the physician hits the bar, they get the reward. This is regardless of the start point. Fortunately for the healthcare organizations yet unfortunately for the already disadvantaged minorities, some physicians and health plans actually get to choose their patients. This way, they could consciously choose to stand at a higher start point by enrolling patients that are easy to care for — or in this case, the non-minorities. In a sense, the pay-for-performance ironically exacerbates the disparities. To make things even worse, this phenomenon is not limited to racial and ethnic disparities; it also manifests in many different aspects of healthcare disparities. One could easily imagine wealthy healthcare organizations enrolling patients with higher socioeconomic status, physicians taking in patients with more manageable symptoms and better prognosis, so and so forth.
A common suggestion to fix the problem is to reward relative improvement as opposed to reaching of an absolute standard. This could be quite promising. But again, as almost all of solutions that people come up with, this is not perfect either. I could see some loop holes in rewarding improvement as well. To put into perspective, some programs use physical data as criteria for rewards. And as you may have imagined, some diseases are intrinsically prone to improvement, whereas some are so easy to go down spiral. Now rewarding improvement essentially brings us back to the problem of “cherry picking” we started with, and creates disparities in the severity of diseases. To ultimately eliminate the disparities in our healthcare system, we still have a very long way to go.
